Determinants of biomass production in hybrid willows and prediction of field performance from pot studies

Abstract

Pot experiments are often performed to assess plant physiological traits and relationships among growth traits under controlled environments. However, the reliability of pot studies for predicting the growth and performance of trees in the field has rarely been rigorously assessed. We evaluated the suitability of pot experiments for predicting field performance, measured as shoot biomass production, by investigating determinants of growth in hybrid willows (Salix spp.) grown under various environmental conditions in the field, and by comparing the data with the results from a corresponding pot study. Biomass production in six hybrid willow clones, bred for use as bio-fuels, was assessed in three field trials located in central and southeastern Sweden throughout the first 3-year cutting cycle. The determinants of biomass productivity, measured as biomass allocation and nitrogen (N) economy, were identified in one of the field trials. Key traits for shoot biomass production in the field were total leaf area and total amount of N; plant N losses by shed leaves were only partly controlled by leaf-litter N concentration. These key traits were also obtained from the pot study and related to shoot biomass production and abscission-leaf N loss in the field. Total leaf area and total N pool of plants grown in pot experiments were good predictors of long-term biomass production in the field, whereas shoot biomass production, specific leaf area and tissue N concentration of pot-grown plants were less suitable as predictors of field performance. Relationships between the key traits and shoot biomass production were clone-specific, indicating the need for analysis of growth traits at the clone level if field performance of trees is to be evaluated based on data from pot studies. Nutrient loss components are important for tree performance in the long term and evaluations of nutrient loss characteristics at the individual-tree level should focus on nutrient pools lost rather than on nutrient concentrations in abscised plant parts.